The majority of firms struggle to get value from their data. A couple of years ago, Forrester reported that between 60% and 73% of the typical company’s data just isn’t used for evaluationThis is because, for technical and security reasons, the info is stored in silos or otherwise pigeonholed, making the appliance of analytics tools difficult – if not inconceivable.
Anna Pojawis and Tyler Maran, engineers who previously worked at Y Combinator-backed startups Hightouch (an information syncing platform) and Fair Square (a medical health insurance tool), were inspired to try their hand at solving the info value problem after noticing that many firms were “locked out” of analytics strategies as a consequence of the technical barriers.
“We found that a significant slice of the market, particularly in regulated industries like healthcare and finance, struggle with data analytics,” Maran told TechCrunch. “Most enterprise data today doesn't slot in a database; it's sales calls, documents, Slack messages, and so forth. And given the scale of those firms, pre-built data models are typically not sufficient.”
So Pojawis and Maran founded OmniAIa set of tools that transform unstructured enterprise data into something that may be understood by data analytics apps and AI.
OmniAI syncs with a corporation's data storage services and databases (e.g. Snowflake, MongoDB, etc.), prepares the info in them, and allows organizations to run the model of their alternative — similar to a big language model — on the info. OmniAI performs all of its work in the corporate's cloud, OmniAI's private cloud, or on-premises environments, and supposedly offers enhanced security, based on Maran.
“We imagine that enormous language models will turn out to be essential to an enterprise’s infrastructure over the following decade and it just is sensible to host all the things in a single place,” Maran said.
OmniAI offers out-of-the-box integrations with models including Meta's Llama 3, Anthropic's Claude, Mistral's Mistral Large, and Amazon's AWS Titan to be used cases similar to routinely redacting sensitive information from data and customarily constructing AI-powered applications. Customers enter right into a software-as-a-service contract with OmniAI to enable management of models on their infrastructure.
It's still early days. But Omni, which recently closed a $3.2 million seed funding round led by FundersClub at a $30 million valuation, claims to have already got 10 customers, including Klaviyo and Carrefour, and annual recurring revenue is predicted to achieve $1 million by 2025, based on Maran.
“We're an incredibly lean team in a fast-growing industry,” Maran said. “We expect that over time, firms will decide to run models alongside their existing infrastructure, and model providers will turn out to be more focused on licensing model weights to existing cloud providers.”
